Logology (science)

Logology is the study of all things related to science and its practitioners—philosophical, biological, psychological, societal, historical, political, institutional, financial. The term "logology" is back-formed from the suffix "-logy", as in "geology", "anthropology", etc., in the sense of the "study of science".^[1][2] The word "logology" provides grammatical variants not available with the earlier terms "science of science" and "sociology of science", such as "logologist", "logologize", "logological", and "logologically".^[a] The emerging field of metascience is a subfield of logology.

The early 20th century brought calls, initially from sociologists, for the creation of a new, empirically based science that would study the scientific enterprise itself.^[5] The early proposals were put forward with some hesitancy and tentativeness.^[6][b] The new meta-science would be given a variety of names,^[8] including "science of knowledge", "science of science", "sociology of science", and "logology".

Florian Znaniecki, who is considered to be the founder of Polish academic sociology, and who in 1954 also served as the 44th president of the American Sociological Association, opened a 1923 article:^[9]

[T]hough theoretical reflection on knowledge—which arose as early as Heraclitus and the Eleatics—stretches... unbroken... through the history of human thought to the present day... we are now witnessing the creation of a new science of knowledge [author's emphasis] whose relation to the old inquiries may be compared with the relation of modern physics and chemistry to the 'natural philosophy' that preceded them, or of contemporary sociology to the 'political philosophy' of antiquity and the Renaissance. [T]here is beginning to take shape a concept of a single, general theory of knowledge... permitting of empirical study.... This theory... is coming to be distinguished clearly from epistemology, from normative logic, and from a strictly descriptive history of knowledge."^[10]

A dozen years later, Polish husband-and-wife sociologists Stanisław Ossowski and Maria Ossowska (the Ossowscy) took up the same subject in an article on "The Science of Science"^[11] whose 1935 English-language version first introduced the term "science of science" to the world.^[12] The article postulated that the new discipline would subsume such earlier ones as epistemology, the philosophy of science, the psychology of science, and the sociology of science.^[13] The science of science would also concern itself with questions of a practical character such as social and state policy in relation to science, such as the organization of institutions of higher learning, of research institutes, and of scientific expeditions, and the protection of scientific workers, etc. It would concern itself as well with historical questions: the history of the conception of science, of the scientist, of the various disciplines, and of learning in general.^[14]

In their 1935 paper, the Ossowscy mentioned the German philosopher Werner Schingnitz (1899–1953) who, in fragmentary 1931 remarks, had enumerated some possible types of research in the science of science and had proposed his own name for the new discipline: scientiology. The Ossowscy took issue with the name:

Those who wish to replace the expression 'science of science' by a one-word term [that] sound[s] international, in the belief that only after receiving such a name [will] a given group of [questions be] officially dubbed an autonomous discipline, [might] be reminded of the name 'mathesiology', proposed long ago for similar purposes [by the French mathematician and physicist André-Marie Ampère (1775–1836)]."^[15]

Yet, before long, in Poland, the unwieldy three-word term nauka o nauce, or science of science, was replaced by the more versatile one-word term naukoznawstwo, or logology, and its natural variants: naukoznawca or logologist, naukoznawczy or logological, and naukoznawczo or logologically. And just after World War II, only 11 years after the Ossowscy's landmark 1935 paper, the year 1946 saw the founding of the Polish Academy of Sciences' quarterly Zagadnienia Naukoznawstwa (Logology) –— long before similar journals in many other countries.^[16][c]

The new discipline also took root elsewhere—in English-speaking countries, without the benefit of a one-word name.

The word science, from the Latin scientia meaning knowledge, signifies somewhat different things in different languages. In English, science when unqualified, generally refers to the exact, natural, or hard sciences.^[18] The corresponding terms in other languages, for example French, German, and Polish, refer to a broader domain that includes not only the exact sciences (logic and mathematics) and the natural sciences (physics, chemistry, biology, Earth sciences, astronomy, etc.) but also the engineering sciences, social sciences (human geography, psychology, cultural anthropology, sociology, political science, economics, linguistics, archaeology, etc.), and humanities (philosophy, history, classics, literary theory, etc.).^[19][d]

University of Amsterdam humanities professor Rens Bod points out that science—defined as a set of methods that describes and interprets observed or inferred phenomena, past or present, aimed at testing hypotheses and building theories—applies to such humanities fields as philology, art history, musicology, philosophy, religious studies, historiography, and literary studies.^[19]

Bod gives a historic example of scientific textual analysis. In 1440 the Italian philologist Lorenzo Valla exposed the Latin document Donatio Constantini, or The Donation of Constantine – which was used by the Catholic Church to legitimize its claim to lands in the Western Roman Empire – as a forgery. Valla used historical, linguistic, and philological evidence, including counterfactual reasoning, to rebut the document. Valla found words and constructions in the document that could not have been used by anyone in the time of Emperor Constantine I, at the beginning of the fourth century C.E. For example, the late Latin word feudum, meaning fief, referred to the feudal system, which would not come into existence until the medieval era, in the seventh century C.E. Valla's methods were those of science, and inspired the later scientifically-minded work of Dutch humanist Erasmus of Rotterdam (1466–1536), Leiden University professor Joseph Justus Scaliger (1540–1609), and philosopher Baruch Spinoza (1632–1677).^[19] Here it is not the experimental method dominant in the exact and natural sciences, but the comparative method central to the humanities, that reigns supreme.

Science's search for the truth about various aspects of reality entails the question of the very knowability of reality. Philosopher Thomas Nagel writes: "[In t]he pursuit of scientific knowledge through the interaction between theory and observation... we test theories against their observational consequences, but we also question or reinterpret our observations in light of theory. (The choice between geocentric and heliocentric theories at the time of the Copernican Revolution is a vivid example.) ... How things seem is the starting point for all knowledge, and its development through further correction, extension, and elaboration is inevitably the result of more seemings—considered judgments about the plausibility and consequences of different theoretical hypotheses. The only way to pursue the truth is to consider what seems true, after careful reflection of a kind appropriate to the subject matter, in light of all the relevant data, principles, and circumstances."^[21]

The question of knowability is approached from a different perspective by physicist-astronomer Marcelo Gleiser: "What we observe is not nature itself but nature as discerned through data we collect from machines. In consequence, the scientific worldview depends on the information we can acquire through our instruments. And given that our tools are limited, our view of the world is necessarily myopic. We can see only so far into the nature of things, and our ever shifting scientific worldview reflects this fundamental limitation on how we perceive reality." Gleiser cites the condition of biology before and after the invention of the microscope or gene sequencing; of astronomy before and after the telescope; of particle physics before and after colliders or fast electronics. "[T]he theories we build and the worldviews we construct change as our tools of exploration transform. This trend is the trademark of science."^[22]

Writes Gleiser: "There is nothing defeatist in understanding the limitations of the scientific approach to knowledge.... What should change is a sense of scientific triumphalism—the belief that no question is beyond the reach of scientific discourse.^{[22] [e]}

"There are clear unknowables in science—reasonable questions that, unless currently accepted laws of nature are violated, we cannot find answers to. One example is the multiverse: the conjecture that our universe is but one among a multitude of others, each potentially with a different set of laws of nature. Other universes lie outside our causal horizon, meaning that we cannot receive or send signals to them. Any evidence for their existence would be circumstantial: for example, scars in the radiation permeating space because of a past collision with a neighboring universe."^[24]

Gleiser gives three further examples of unknowables, involving the origins of the universe; of life; and of mind:^[24][f]

"Scientific accounts of the origin of the universe are incomplete because they must rely on a conceptual framework to even begin to work: energy conservation, relativity, quantum physics, for instance. Why does the universe operate under these laws and not others?^[24]

"Similarly, unless we can prove that only one or very few biochemical pathways exist from nonlife to life, we cannot know for sure how life originated on Earth.^[24]

"For consciousness, the problem is the jump from the material to the subjective—for example, from firing neurons to the experience of pain or the color red. Perhaps some kind of rudimentary consciousness could emerge in a sufficiently complex machine. But how could we tell? How do we establish—as opposed to conjecture—that something is conscious?"^[24] Paradoxically, writes Gleiser, it is through our consciousness that we make sense of the world, even if imperfectly. "Can we fully understand something of which we are a part?"^[24]

Among all the sciences (i.e., disciplines of learning, writ large) there seems to exist an inverse relation between precision and intuitiveness. The most intuitive of the disciplines, aptly termed the "humanities", relate to common human experience and, even at their most exact, are thrown back on the comparative method; less intuitive and more precise than the humanities are the social sciences; while, at the base of the inverted pyramid of the disciplines, physics (concerned with mattergy – the matter and energy comprising the universe) is, at its deepest, the most precise discipline and at the same time utterly non-intuitive.^[g][h]

Theoretical physicist and mathematician Freeman Dyson explains that "[s]cience consists of facts and theories":

"Facts are supposed to be true or false. They are discovered by observers or experimenters. A scientist who claims to have discovered a fact that turns out to be wrong is judged harshly....

"Theories have an entirely different status. They are free creations of the human mind, intended to describe our understanding of nature. Since our understanding is incomplete, theories are provisional. Theories are tools of understanding, and a tool does not need to be precisely true in order to be useful. Theories are supposed to be more-or-less true... A scientist who invents a theory that turns out to be wrong is judged leniently."^[26]

Dyson cites a psychologist's description of how theories are born: "We can't live in a state of perpetual doubt, so we make up the best story possible and we live as if the story were true." Dyson writes: "The inventor of a brilliant idea cannot tell whether it is right or wrong." The passionate pursuit of wrong theories is a normal part of the development of science.^[27] Dyson cites, after Mario Livio, five famous scientists who made major contributions to the understanding of nature but also believed firmly in a theory that proved wrong.^[27]

Charles Darwin explained the evolution of life with his theory of natural selection of inherited variations, but he believed in a theory of blending inheritance that made the propagation of new variations impossible.^[27] He never read Gregor Mendel's studies that showed that the laws of inheritance would become simple when inheritance was considered as a random process. Though Darwin in 1866 did the same experiment that Mendel had, Darwin did not get comparable results because he failed to appreciate the statistical importance of using very large experimental samples. Eventually, Mendelian inheritance by random variation would, no thanks to Darwin, provide the raw material for Darwinian selection to work on.^[28]

William Thomson (Lord Kelvin) discovered basic laws of energy and heat, then used these laws to calculate an estimate of the age of the Earth that was too short by a factor of fifty. He based his calculation on the belief that the Earth's mantle was solid and could transfer heat from the interior to the surface only by conduction. It is now known that the mantle is partly fluid and transfers most of the heat by the far more efficient process of convection, which carries heat by a massive circulation of hot rock moving upward and cooler rock moving downward. Kelvin could see the eruptions of volcanoes bringing hot liquid from deep underground to the surface; but his skill in calculation blinded him to processes, such as volcanic eruptions, that could not be calculated.^[27]

Linus Pauling discovered the chemical structure of protein and proposed a completely wrong structure for DNA, which carries hereditary information from parent to offspring. Pauling guessed a wrong structure for DNA because he assumed that a pattern that worked for protein would also work for DNA. He overlooked the gross chemical differences between protein and DNA. Francis Crick and James Watson paid attention to the differences and found the correct structure for DNA that Pauling had missed a year earlier.^[27]

Astronomer Fred Hoyle discovered the process by which the heavier elements essential to life are created by nuclear reactions in the cores of massive stars. He then proposed a theory of the history of the universe known as steady-state cosmology, which has the universe existing forever without an initial Big Bang (as Hoyle derisively dubbed it). He held his belief in the steady state long after observations proved that the Big Bang had happened.^[27]

Albert Einstein discovered the theory of space, time, and gravitation known as general relativity, and then added a cosmological constant, later known as dark energy. Subsequently, Einstein withdrew his proposal of dark energy, believing it unnecessary. Long after his death, observations suggested that dark energy really exists, so that Einstein's addition to the theory may have been right; and his withdrawal, wrong.^[27]

To Mario Livio's five examples of scientists who blundered, Dyson adds a sixth: himself. Dyson had concluded, on theoretical principles, that what was to become known as the W-particle, a charged weak boson, could not exist. An experiment conducted at CERN, in Geneva, later proved him wrong. "With hindsight I could see several reasons why my stability argument would not apply to W-particles. [They] are too massive and too short-lived to be a constituent of anything that resembles ordinary matter."^[29]

Harvard University historian of science Naomi Oreskes points out that the truth of scientific findings can never be assumed to be finally, absolutely settled.^[30] The history of science offers many examples of matters that scientists once thought to be settled and which have proven not to be, such as the concepts of Earth being the center of the universe, the absolute nature of time and space, the stability of continents, and the cause of infectious disease.^[30]

Science, writes Oreskes, is not a fixed, immutable set of discoveries but "a process of learning and discovery [...]. Science can also be understood as an institution (or better, a set of institutions) that facilitates this work.^[30]

It is often asserted that scientific findings are true because scientists use "the scientific method". But, writes Oreskes, "we can never actually agree on what that method is. Some will say it is empiricism: observation and description of the world. Others will say it is the experimental method: the use of experience and experiment to test hypotheses. (This is cast sometimes as the hypothetico-deductive method, in which the experiment must be framed as a deduction from theory, and sometimes as falsification, where the point of observation and experiment is to refute theories, not to confirm them.) Recently a prominent scientist claimed the scientific method was to avoid fooling oneself into thinking something is true that is not, and vice versa."^[30]

In fact, writes Oreskes, the methods of science have varied between disciplines and across time. "Many scientific practices, particularly statistical tests of significance, have been developed with the idea of avoiding wishful thinking and self-deception, but that hardly constitutes 'the scientific method.'"^[30]

Science, writes Oreskes, "is not simple, and neither is the natural world; therein lies the challenge of science communication. [...] Our efforts to understand and characterize the natural world are just that: efforts. Because we're human, we often fall flat."^[30]

"Scientific theories", according to Oreskes, "are not perfect replicas of reality, but we have good reason to believe that they capture significant elements of it."^[30]

Steven Weinberg, 1979 Nobel laureate in physics, and a historian of science, writes that the core goal of science has always been the same: "to explain the world"; and in reviewing earlier periods of scientific thought, he concludes that only since Isaac Newton has that goal been pursued more or less correctly. He decries the "intellectual snobbery" that Plato and Aristotle showed in their disdain for science's practical applications, and he holds Francis Bacon and René Descartes to have been the "most overrated" among the forerunners of modern science (they tried to prescribe rules for conducting science, which "never works").^[31]

Weinberg draws parallels between past and present science, as when a scientific theory is "fine-tuned" (adjusted) to make certain quantities equal, without any understanding of why they should be equal. Such adjusting vitiated the celestial models of Plato's followers, in which different spheres carrying the planets and stars were assumed, with no good reason, to rotate in exact unison. But, Weinberg writes, a similar fine-tuning also besets current efforts to understand the "dark energy" that is speeding up the expansion of the universe.^[32]

Ancient science has been described as having gotten off to a good start, then faltered. The doctrine of atomism, propounded by the pre-Socratic philosophers Leucippus and Democritus, was naturalistic, accounting for the workings of the world by impersonal processes, not by divine volitions. Nevertheless, these pre-Socratics come up short for Weinberg as proto-scientists, in that they apparently never tried to justify their speculations or to test them against evidence.^[32]

Weinberg believes that science faltered early on due to Plato's suggestion that scientific truth could be attained by reason alone, disregarding empirical observation, and due to Aristotle's attempt to explain nature teleologically—in terms of ends and purposes. Plato's ideal of attaining knowledge of the world by unaided intellect was "a false goal inspired by mathematics"—one that for centuries "stood in the way of progress that could be based only on careful analysis of careful observation." And it "never was fruitful" to ask, as Aristotle did, "what is the purpose of this or that physical phenomenon."^[32]

A scientific field in which the Greek and Hellenistic world did make progress was astronomy. This was partly for practical reasons: the sky had long served as compass, clock, and calendar. Also, the regularity of the movements of heavenly bodies made them simpler to describe than earthly phenomena. But not too simple: though the sun, moon and "fixed stars" seemed regular in their celestial circuits, the "wandering stars"—the planets—were puzzling; they seemed to move at variable speeds, and even to reverse direction. Writes Weinberg: "Much of the story of the emergence of modern science deals with the effort, extending over two millennia, to explain the peculiar motions of the planets."^[33]

The challenge was to make sense of the apparently irregular wanderings of the planets on the assumption that all heavenly motion is actually circular and uniform in speed. Circular, because Plato held the circle to be the most perfect and symmetrical form; and therefore circular motion, at uniform speed, was most fitting for celestial bodies. Aristotle agreed with Plato. In Aristotle's cosmos, everything had a "natural" tendency to motion that fulfilled its inner potential. For the cosmos' sublunary part (the region below the Moon), the natural tendency was to move in a straight line: downward, for earthen things (such as rocks) and water; upward, for air and fiery things (such as sparks). But in the celestial realm things were not composed of earth, water, air, or fire, but of a "fifth element", or "quintessence," which was perfect and eternal. And its natural motion was uniformly circular. The stars, the Sun, the Moon, and the planets were carried in their orbits by a complicated arrangement of crystalline spheres, all centered around an immobile Earth.^[34]

The Platonic-Aristotelian conviction that celestial motions must be circular persisted stubbornly. It was fundamental to the astronomer Ptolemy's system, which improved on Aristotle's in conforming to the astronomical data by allowing the planets to move in combinations of circles called "epicycles".^[34]

It even survived the Copernican Revolution. Copernicus was conservative in his Platonic reverence for the circle as the heavenly pattern. According to Weinberg, Copernicus was motivated to dethrone the Earth in favor of the Sun as the immobile center of the cosmos largely by aesthetic considerations: he objected to the fact that Ptolemy, though faithful to Plato's requirement that heavenly motion be circular, had departed from Plato's other requirement that it be of uniform speed. By putting the sun at the center—actually, somewhat off-center—Copernicus sought to honor circularity while restoring uniformity. But to make his system fit the observations as well as Ptolemy's system, Copernicus had to introduce still more epicycles. That was a mistake that, writes Weinberg, illustrates a recurrent theme in the history of science: "A simple and beautiful theory that agrees pretty well with observation is often closer to the truth than a complicated ugly theory that agrees better with observation."^[34]

The planets, however, do not move in perfect circles but in ellipses. It was Johannes Kepler, about a century after Copernicus, who reluctantly (for he too had Platonic affinities) realized this. Thanks to his examination of the meticulous observations compiled by astronomer Tycho Brahe, Kepler "was the first to understand the nature of the departures from uniform circular motion that had puzzled astronomers since the time of Plato."^[34]

The replacement of circles by supposedly ugly ellipses overthrew Plato's notion of perfection as the celestial explanatory principle. It also destroyed Aristotle's model of the planets carried in their orbits by crystalline spheres; writes Weinberg, "there is no solid body whose rotation can produce an ellipse." Even if a planet were attached to an ellipsoid crystal, that crystal's rotation would still trace a circle. And if the planets were pursuing their elliptical motion through empty space, then what was holding them in their orbits?^[34]

Science had reached the threshold of explaining the world not geometrically, according to shape, but dynamically, according to force. It was Isaac Newton who finally crossed that threshold. He was the first to formulate, in his "laws of motion", the concept of force. He demonstrated that Kepler's ellipses were the very orbits the planets would take if they were attracted toward the Sun by a force that decreased as the square of the planet's distance from the Sun. And by comparing the Moon's motion in its orbit around the Earth to the motion of, perhaps, an apple as it falls to the ground, Newton deduced that the forces governing them were quantitatively the same. "This," writes Weinberg, "was the climactic step in the unification of the celestial and terrestrial in science."^[34]

By formulating a unified explanation of the behavior of planets, comets, moons, tides, and apples, writes Weinberg, Newton "provided an irresistible model for what a physical theory should be"—a model that fit no preexisting metaphysical criterion. In contrast to Aristotle, who claimed to explain the falling of a rock by appeal to its inner striving, Newton was unconcerned with finding a deeper cause for gravity.^[34] He declared in a postscript to the second, 1713 edition of his Philosophiæ Naturalis Principia Mathematica: "I have not as yet been able to deduce from phenomena the reason for these properties of gravity, and I do not feign hypotheses. It is enough that gravity really exists and acts according to the laws that we have set forth."^[35] What mattered were his mathematically stated principles describing this force, and their ability to account for a vast range of phenomena.^[34]

About two centuries later, in 1915, a deeper explanation for Newton's law of gravitation was found in Albert Einstein's general theory of relativity: gravity could be explained as a manifestation of the curvature in spacetime resulting from the presence of matter and energy. Successful theories like Newton's, writes Weinberg, may work for reasons that their creators do not understand—reasons that deeper theories will later reveal. Scientific progress is not a matter of building theories on a foundation of reason, but of unifying a greater range of phenomena under simpler and more general principles.^[34]

Naomi Oreskes cautions against making "the classic error of conflating absence of evidence with evidence of absence [emphases added]." She cites two examples of this error that were perpetrated in 2016 and 2023.^[36]

In 2016 the Cochrane Library, a collection of databases in medicine and other healthcare specialties, published a report that was widely understood to indicate that flossing one's teeth confers no advantage to dental health. But the American Academy of Periodontology, dental professors, deans of dental schools, and clinical dentists all held that clinical practice shows differences in tooth and gum health between those who floss and those who don't.^[37]

Oreskes explains that "Cochrane Reviews base their findings on randomized controlled trials (RCTs), often called the 'gold standard' of scientific evidence." But many questions can't be answered well using this method, and some can't be answered at all. "Nutrition is a case in point. [Y]ou can't control what people eat, and when you ask... what they have eaten, many people lie. Flossing is similar. One survey concluded that one in four Americans who claimed to floss regularly was fibbing."^[38]

In 2023 Cochrane published a report determining that wearing surgical masks "probably makes little or no difference" in slowing the spread of respiratory illnesses such as COVID-19. Mass media reduced this to the claim that masks did not work. The Cochrane Library's editor-in-chief objected to such characterizations of the review; she said the report had not concluded that "masks don't work", but rather that the "results were inconclusive." The report had made clear that its conclusions were about the quality and capaciousness of available evidence, which the authors felt were insufficient to prove that masking was effective. The report's authors were "uncertain whether wearing [surgical] masks or N95/P2 respirators helps to slow the spread of respiratory viruses." Still, they were also uncertain about that uncertainty [emphasis added], stating that their confidence in their conclusion was "low to moderate."^[39]

Subsequently the report's lead author confused the public by stating that mask-wearing "Makes no difference – none of it", and that Covid policies were "evidence-free": he thus perpetrated what Oreskes calls "the [...] error of conflating absence of evidence with evidence of absence." Studies have in fact shown that U.S. states with mask mandates saw a substantial decline in Covid spread within days of mandate orders being signed; in the period from 31 March to 22 May 2020, more than 200,000 cases were avoided.^[40]

Oreskes calls the Cochrane report's neglect of the epidemiological evidence – because it didn't meet Cochrane's rigid standard – "methodological fetishism," when scientists "fixate on a preferred methodology and dismiss studies that don't follow it."^[41]

The term "artificial intelligence" (AI) was coined in 1955 by John McCarthy when he and other computer scientists were planning a workshop and did not want to invite Norbert Wiener, the brilliant, pugnacious, and increasingly philosophical (rather than practical) author on feedback mechanisms who had coined the term "cybernetics". The new term artificial intelligence, writes Kenneth Cukier, "set in motion decades of semantic squabbles ('Can machines think?') and fueled anxieties over malicious robots... If McCarthy... had chosen a blander phrase—say, 'automation studies'—the concept might not have appealed as much to Hollywood [movie] producers and [to] journalists..."^[42] Similarly Naomi Oreskes has commented: "[M]achine 'intelligence'... isn't intelligence at all but something more like 'machine capability.'"^[43]

As machines have become increasingly capable, specific tasks considered to require "intelligence", such as optical character recognition, have often been removed from the definition of AI, a phenomenon known as the "AI effect". It has been quipped that "AI is whatever hasn't been done yet."^[44]

Since 1950, when Alan Turing proposed what has come to be called the "Turing test," there has been speculation whether machines such as computers can possess intelligence; and, if so, whether intelligent machines could become a threat to human intellectual and scientific ascendancy—or even an existential threat to humanity.^[45] John Searle points out common confusion about the correct interpretation of computation and information technology. "For example, one routinely reads that in exactly the same sense in which Garry Kasparov… beat Anatoly Karpov in chess, the computer called Deep Blue played and beat Kasparov.... [T]his claim is [obviously] suspect. In order for Kasparov to play and win, he has to be conscious that he is playing chess, and conscious of a thousand other things... Deep Blue is conscious of none of these things because it is not conscious of anything at all. Why is consciousness so important? You cannot literally play chess or do much of anything else cognitive if you are totally disassociated from consciousness."^[45]

Searle explains that, "in the literal, real, observer-independent sense in which humans compute, mechanical computers do not compute. They go through a set of transitions in electronic states that we can interpret computationally. The transitions in those electronic states are absolute or observer-independent, but the computation is observer-relative. The transitions in physical states are just electrical sequences unless some conscious agent can give them a computational interpretation.... There is no psychological reality at all to what is happening in the [computer]."^[46]

"[A] digital computer", writes Searle, "is a syntactical machine. It manipulates symbols and does nothing else. For this reason, the project of creating human intelligence by designing a computer program that will pass the Turing Test... is doomed from the start. The appropriately programmed computer has a syntax [rules for constructing or transforming the symbols and words of a language] but no semantics [comprehension of meaning].... Minds, on the other hand, have mental or semantic content."^[47]

Like Searle, Christof Koch, chief scientist and president of the Allen Institute for Brain Science, in Seattle, is doubtful about the possibility of "intelligent" machines attaining consciousness, because "[e]ven the most sophisticated brain simulations are unlikely to produce conscious feelings." According to Koch, "Whether machines can become sentient [is important] for ethical reasons. If computers experience life through their own senses, they cease to be purely a means to an end determined by their usefulness to... humans. Per GNW [the Global Neuronal Workspace theory], they turn from mere objects into subjects... with a point of view.... Once computers' cognitive abilities rival those of humanity, their impulse to push for legal and political rights will become irresistible – the right not to be deleted, not to have their memories wiped clean, not to suffer pain and degradation. The alternative, embodied by IIT [Integrated Information Theory], is that computers will remain only supersophisticated machinery, ghostlike empty shells, devoid of what we value most: the feeling of life itself."^[48]

Professor of psychology and neural science Gary Marcus points out a so far insuperable stumbling block to artificial intelligence: an incapacity for reliable disambiguation. "[V]irtually every sentence [that people generate] is ambiguous, often in multiple ways. Our brain is so good at comprehending language that we do not usually notice."^[49] A prominent example is known as the "pronoun disambiguation problem" ("PDP"): a machine has no way of determining to whom or what a pronoun in a sentence—such as "he", "she" or "it"—refers.^[50]

Marcus has described current large language models as "approximations to [...] language use rather than language understanding".^[51]

Computer scientist Pedro Domingos writes: "AIs are like autistic savants and will remain so for the foreseeable future.... AIs lack common sense and can easily make errors that a human never would... They are also liable to take our instructions too literally, giving us precisely what we asked for instead of what we actually wanted.^[52]

Kai-Fu Lee, a Beijing-based venture capitalist, artificial-intelligence (AI) expert with a Ph.D. in computer science from Carnegie Mellon University, and author of the 2018 book, AI Superpowers: China, Silicon Valley, and the New World Order,^[53] emphasized in a 2018 PBS Amanpour interview with Hari Sreenivasan that AI, with all its capabilities, will never be capable of creativity or empathy.^[54] Paul Scharre writes in Foreign Affairs that "Today's AI technologies are powerful but unreliable."^[55][i] George Dyson, historian of computing, writes (in what might be called "Dyson's Law") that "Any system simple enough to be understandable will not be complicated enough to behave intelligently, while any system complicated enough to behave intelligently will be too complicated to understand."^[57] Computer scientist Alex Pentland writes: "Current AI machine-learning algorithms are, at their core, dead simple stupid. They work, but they work by brute force."^[58]

"Artificial intelligence" is synonymous with "machine intelligence." The more perfectly adapted an AI program is to a given task, the less applicable it will be to other specific tasks. An abstracted, AI general intelligence is a remote prospect, if feasible at all. Melanie Mitchell notes that an AI program called AlphaGo bested one of the world's best Go players, but that its "intelligence" is nontransferable: it cannot "think" about anything except Go. Mitchell writes: "We humans tend to overestimate AI advances and underestimate the complexity of our own intelligence."^[59] Writes Paul Taylor: "Perhaps there is a limit to what a computer can do without knowing that it is manipulating imperfect representations of an external reality."^[60]

Humankind may not be able to outsource, to machines, its creative efforts in the sciences, technology, and culture.

Gary Marcus cautions against being taken in by deceptive claims about artificial general intelligence capabilities that are put out in press releases by self-interested companies which tell the press and public "only what the companies want us to know."^[61] Marcus writes:

Although deep learning has advanced the ability of machines to recognize patterns in data, it has three major flaws. The patterns that it learns are, ironically, superficial not conceptual; the results it creates are hard to interpret; and the results are difficult to use in the context of other processes, such as memory and reasoning. As Harvard University computer scientist Les Valiant noted, "The central challenge [going forward] is to unify the formulation of... learning and reasoning."^[62]

James Gleick writes: "Agency is what distinguishes us from machines. For biological creatures, reason and purpose come from acting in the world and experiencing the consequences. Artificial intelligences – disembodied, strangers to blood, sweat, and tears – have no occasion for that."^[63]

A central concern for science and scholarship is the reliability and reproducibility of their findings. Of all fields of study, none is capable of such precision as physics. But even there the results of studies, observations, and experiments cannot be considered absolutely certain and must be treated probabilistically; hence, statistically.^[64]

In 1925 British geneticist and statistician Ronald Fisher published Statistical Methods for Research Workers, which established him as the father of modern statistics. He proposed a statistical test that summarized the compatibility of data with a given proposed model and produced a "p value". He counselled pursuing results with p values below 0.05 and not wasting time on results above that. Thus arose the idea that a p value less than 0.05 constitutes "statistical significance" – a mathematical definition of "significant" results.^[65]

The use of p values, ever since, to determine the statistical significance of experimental results has contributed to an illusion of certainty and to reproducibility crises in many scientific fields,^[66] especially in experimental economics, biomedical research, and psychology.^[67]

Every statistical model relies on a set of assumptions about how data are collected and analyzed and about how researchers decide to present their results. These results almost always center on null-hypothesis significance testing, which produces a p value. Such testing does not address the truth head-on but obliquely: significance testing is meant to indicate only whether a given line of research is worth pursuing further. It does not say how likely the hypothesis is to be true, but instead addresses an alternative question: if the hypothesis were false, how unlikely would the data be? The importance of "statistical significance", reflected in the p value, can be exaggerated or overemphasized – something that readily occurs with small samples. That has caused replication crises.^[64]

Some scientists have advocated "redefining statistical significance", shifting its threshold from 0.05 to 0.005 for claims of new discoveries. Others say such redefining does no good because the real problem is the very existence of a threshold.^[68]

Some scientists prefer to use Bayesian methods, a more direct statistical approach which takes initial beliefs, adds in new evidence, and updates the beliefs. Another alternative procedure is to use the surprisal, a mathematical quantity that adjust p values to produce bits – as in computer bits – of information; in that perspective, 0.05 is a weak standard.^[68]

When Ronald Fisher embraced the concept of "significance" in the early 20th century, it meant "signifying" but not "important". Statistical "significance" has, since, acquired am excessive connotation of confidence in the validity of the experimental results. Statistician Andrew Gelman says, "The original sin is people wanting certainty when it's not appropriate." "Ultimately", writes Lydia Denworth, "a successful theory is one that stands up repeatedly to decades of scrutiny."^[68]

Increasingly, attention is being given to the principles of open science, such as publishing more detailed research protocols and requiring authors to follow prespecified analysis plans and to report when they deviate from them.^[68]

Fifty years before Florian Znaniecki published his 1923 paper proposing the creation of an empirical field of study to study the field of science, Aleksander Głowacki (better known by his pen name, Bolesław Prus) had made the same proposal. In an 1873 public lecture "On Discoveries and Inventions",^[69] Prus said:

Until now there has been no science that describes the means for making discoveries and inventions, and the generality of people, as well as many men of learning, believe that there never will be. This is an error. Someday a science of making discoveries and inventions will exist and will render services. It will arise not all at once; first only its general outline will appear, which subsequent researchers will correct and elaborate, and which still later researchers will apply to individual branches of knowledge.^[70]

Prus defines "discovery" as "the finding out of a thing that has existed and exists in nature, but which was previously unknown to people";^[71] and "invention" as "the making of a thing that has not previously existed, and which nature itself cannot make."^[72]

He illustrates the concept of "discovery":

Until 400 years ago, people thought that the Earth comprised just three parts: Europe, Asia, and Africa; it was only in 1492 that the Genoese, Christopher Columbus, sailed out from Europe into the Atlantic Ocean and, proceeding ever westward, after [10 weeks] reached a part of the world that Europeans had never known. In that new land he found copper-colored people who went about naked, and he found plants and animals different from those in Europe; in short, he had discovered a new part of the world that others would later name "America." We say that Columbus had discovered America, because America had already long existed on Earth.^[73]

Prus illustrates the concept of "invention":

[As late as] 50 years ago, locomotives were unknown, and no one knew how to build one; it was only in 1828 that the English engineer Stephenson built the first locomotive and set it in motion. So we say that Stephenson invented the locomotive, because this machine had not previously existed and could not by itself have come into being in nature; it could only have been made by man.^[72]

According to Prus, "inventions and discoveries are natural phenomena and, as such, are subject to certain laws." Those are the laws of "gradualness", "dependence", and "combination".^[74]

1. The law of gradualness. No discovery or invention arises at once perfected, but is perfected gradually; likewise, no invention or discovery is the work of a single individual but of many individuals, each adding his little contribution.^[75]

2. The law of dependence. An invention or discovery is conditional on the prior existence of certain known discoveries and inventions. ...If the rings of Saturn can [only] be seen through telescopes, then the telescope had to have been invented before the rings could have been seen. [...]^[76]

3. The law of combination. Any new discovery or invention is a combination of earlier discoveries and inventions, or rests on them. When I study a new mineral, I inspect it, I smell it, I taste it ... I combine the mineral with a balance and with fire...in this way I learn ever more of its properties.^[77][j]

Each of Prus' three "laws" entails important corollaries. The law of gradualness implies the following:^[79]

a) Since every discovery and invention requires perfecting, let us not pride ourselves only on discovering or inventing something completely new, but let us also work to improve or get to know more exactly things that are already known and already exist. [...]^[79] b) The same law of gradualness demonstrates the necessity of expert training. Who can perfect a watch, if not a watchmaker with a good comprehensive knowledge of his métier? Who can discover new characteristics of an animal, if not a naturalist?^[79]

From the law of dependence flow the following corollaries:^[79]

a) No invention or discovery, even one seemingly without value, should be dismissed, because that particular trifle may later prove very useful. There would seem to be no simpler invention than the needle, yet the clothing of millions of people, and the livelihoods of millions of seamstresses, depend on the needle's existence. Even today's beautiful sewing machine would not exist, had the needle not long ago been invented.^[80] b) The law of dependence teaches us that what cannot be done today, might be done later. People give much thought to the construction of a flying machine that could carry many persons and parcels. The inventing of such a machine will depend, among other things, on inventing a material that is, say, as light as paper and as sturdy and fire-resistant as steel.^[81]

Finally, Prus' corollaries to his law of combination:^[81]

a) Anyone who wants to be a successful inventor, needs to know a great many things—in the most diverse fields. For if a new invention is a combination of earlier inventions, then the inventor's mind is the ground on which, for the first time, various seemingly unrelated things combine. Example: The steam engine combines the kettle for cooking Rumford's Soup, the pump, and the spinning wheel.^[81]

[...] What is the connection among zinc, copper, sulfuric acid, a magnet, a clock mechanism, and an urgent message? All these had to come together in the mind of the inventor of the telegraph... [...]^[82]

The greater the number of inventions that come into being, the more things a new inventor must know; the first, earliest and simplest inventions were made by completely uneducated people—but today's inventions, particularly scientific ones, are products of the most highly educated minds. [...]^[83]

b) A second corollary concerns societies that wish to have inventors. I said that a new invention is created by combining the most diverse objects; let us see where this takes us.^[83]

Suppose I want to make an invention, and someone tells me: Take 100 different objects and bring them into contact with one another, first two at a time, then three at a time, finally four at a time, and you will arrive at a new invention. Imagine that I take a burning candle, charcoal, water, paper, zinc, sugar, sulfuric acid, and so on, 100 objects in all, and combine them with one another, that is, bring into contact first two at a time: charcoal with flame, water with flame, sugar with flame, zinc with flame, sugar with water, etc. Each time, I shall see a phenomenon: thus, in fire, sugar will melt, charcoal will burn, zinc will heat up, and so on. Now I will bring into contact three objects at a time, for example, sugar, zinc and flame; charcoal, sugar and flame; sulfuric acid, zinc and water; etc., and again I shall experience phenomena. Finally I bring into contact four objects at a time, for example, sugar, zinc, charcoal, and sulfuric acid. Ostensibly this is a very simple method, because in this fashion I could make not merely one but a dozen inventions. But will such an effort not exceed my capability? It certainly will. A hundred objects, combined in twos, threes and fours, will make over 4 million combinations; so if I made 100 combinations a day, it would take me over 110 years to exhaust them all!^[84]

But if by myself I am not up to the task, a sizable group of people will be. If 1,000 of us came together to produce the combinations that I have described, then any one person would only have to carry out slightly more than 4,000 combinations. If each of us performed just 10 combinations a day, together we would finish them all in less than a year and a half: 1,000 people would make an invention which a single man would have to spend more than 110 years to make…^[85][k]

The conclusion is quite clear: a society that wants to win renown with its discoveries and inventions has to have a great many persons working in every branch of knowledge. One or a few men of learning and genius mean nothing today, or nearly nothing, because everything is now done by large numbers. I would like to offer the following simile: Inventions and discoveries are like a lottery; not every player wins, but from among the many players a few must win. The point is not that John or Paul, because they want to make an invention and because they work for it, shall make an invention; but where thousands want an invention and work for it, the invention must appear, as surely as an unsupported rock must fall to the ground.^[85][l]

But, asks Prus, "What force drives [the] toilsome, often frustrated efforts [of the investigators]? What thread will clew these people through hitherto unexplored fields of study?"^[86][m]

[T]he answer is very simple: man is driven to efforts, including those of making discoveries and inventions, by needs; and the thread that guides him is observation: observation of the works of nature and of man.^[86]

I have said that the mainspring of all discoveries and inventions is needs. In fact, is there any work of man that does not satisfy some need? We build railroads because we need rapid transportation; we build clocks because we need to measure time; we build sewing machines because the speed of [unaided] human hands is insufficient. We abandon home and family and depart for distant lands because we are drawn by curiosity to see what lies elsewhere. We forsake the society of people and we spend long hours in exhausting contemplation because we are driven by a hunger for knowledge, by a desire to solve the challenges that are constantly thrown up by the world and by life!^[86]

Needs never cease; on the contrary, they are always growing. While the pauper thinks about a piece of bread for lunch, the rich man thinks about wine after lunch. The foot traveler dreams of a rudimentary wagon; the railroad passenger demands a heater. The infant is cramped in its cradle; the mature man is cramped in the world. In short, everyone has his needs, and everyone desires to satisfy them, and that desire is an inexhaustible source of new discoveries, new inventions, in short, of all progress.^[87]

But needs are general, such as the needs for food, sleep and clothing; and special, such as needs for a new steam engine, a new telescope, a new hammer, a new wrench. To understand the former needs, it suffices to be a human being; to understand the latter needs, one must be a specialist—an expert worker. Who knows better than a tailor what it is that tailors need, and who better than a tailor knows how to find the right way to satisfy the need?^[88]

Now consider how observation can lead man to new ideas; and to that end, as an example, let us imagine how, more or less, clay products came to be invented.^[88]

Suppose that somewhere there lived on clayey soil a primitive people who already knew fire. When rain fell on the ground, the clay turned doughy; and if, shortly after the rain, a fire was set on top of the clay, the clay under the fire became fired and hardened. If such an event occurred several times, the people might observe and thereafter remember that fired clay becomes hard like stone and does not soften in water. One of the primitives might also, when walking on wet clay, have impressed deep tracks into it; after the sun had dried the ground and rain had fallen again, the primitives might have observed that water remains in those hollows longer than on the surface. Inspecting the wet clay, the people might have observed that this material can be easily kneaded in one's fingers and accepts various forms.^[89]

Some ingenious persons might have started shaping clay into various animal forms [...] etc., including something shaped like a tortoise shell, which was in use at the time. Others, remembering that clay hardens in fire, might have fired the hollowed-out mass, thereby creating the first [clay] bowl.^[90]

After that, it was a relatively easy matter to perfect the new invention; someone else could discover clay more suitable for such manufactures; someone else could invent a glaze, and so on, with nature and observation at every step pointing out to man the way to invention. [...]^[90]

[This example] illustrates how people arrive at various ideas: by closely observing all things and wondering about all things.^[90]

Take another example. [S]ometimes, in a pane of glass, we find disks and bubbles, looking through which we see objects more distinctly than with the naked eye. Suppose that an alert person, spotting such a bubble in a pane, took out a piece of glass and showed it to others as a toy. Possibly among them there was a man with weak vision who found that, through the bubble in the pane, he saw better than with the naked eye. Closer investigation showed that bilaterally convex glass strengthens weak vision, and in this way eyeglasses were invented. People may first have cut glass for eyeglasses from glass panes, but in time others began grinding smooth pieces of glass into convex lenses and producing proper eyeglasses.^[91]

The art of grinding eyeglasses was known almost 600 years ago. A couple of hundred years later, the children of a certain eyeglass grinder, while playing with lenses, placed one in front of another and found that they could see better through two lenses than through one. They informed their father about this curious occurrence, and he began producing tubes with two magnifying lenses and selling them as a toy. Galileo, the great Italian scientist, on learning of this toy, used it for a different purpose and built the first telescope.^[92]

This example, too, shows us that observation leads man by the hand to inventions. This example again demonstrates the truth of gradualness in the development of inventions, but above all also the fact that education amplifies man's inventiveness. A simple lens-grinder formed two magnifying glasses into a toy—while Galileo, one of the most learned men of his time, made a telescope. As Galileo's mind was superior to the craftsman's mind, so the invention of the telescope was superior to the invention of a toy.^[92] [...]

The three laws [that have been discussed here] are immensely important and do not apply only to discoveries and inventions, but they pervade all of nature. An oak does not immediately become an oak but begins as an acorn, then becomes a seedling, later a little tree, and finally a mighty oak: we see here the law of gradualness. A seed that has been sown will not germinate until it finds sufficient heat, water, soil and air: here we see the law of dependence. Finally, no animal or plant, or even stone, is something homogeneous and simple but is composed of various organs: here we see the law of combination.^[93]

Prus holds that, over time, the multiplication of discoveries and inventions has improved the quality of people's lives and has expanded their knowledge. "This gradual advance of civilized societies, this constant growth in knowledge of the objects that exist in nature, this constant increase in the number of tools and useful materials, is termed progress, or the growth of civilization."^[94] Conversely, Prus warns, "societies and people that do not make inventions or know how to use them, lead miserable lives and ultimately perish."^[95][n]

A fundamental feature of the scientific enterprise is reproducibility of results. "For decades", writes Shannon Palus, "it has been... an open secret that a [considerable part] of the literature in some fields is plain wrong." This effectively sabotages the scientific enterprise and costs the world many billions of dollars annually in wasted resources. Militating against reproducibility is scientists' reluctance to share techniques, for fear of forfeiting one's advantage to other scientists. Also, scientific journals and tenure committees tend to prize impressive new results rather than gradual advances that systematically build on existing literature. Scientists who quietly fact-check others' work or spend extra time ensuring that their own protocols are easy for other researchers to understand, gain little for themselves.^[96]

With a view to improving reproducibility of scientific results, it has been suggested that research-funding agencies finance only projects that include a plan for making their work transparent. In 2016 the U.S. National Institutes of Health introduced new application instructions and review questions to encourage scientists to improve reproducibility. The NIH requests more information on how the study builds on previous work, and a list of variables that could affect the study, such as the sex of animal subjects—a previously overlooked factor that led many studies to describe phenomena found in male animals as universal.^[97]

Likewise, the questions that a funder can ask in advance could be asked by journals and reviewers. One solution is "registered reports", a preregistration of studies whereby a scientist submits, for publication, research analysis and design plans before actually doing the study. Peer reviewers then evaluate the methodology, and the journal promises to print the results, no matter what they are. In order to prevent over-reliance on preregistered studies—which could encourage safer, less venturesome research, thus over-correcting the problem—the preregistered-studies model could be operated in tandem with the traditional results-focused model, which may sometimes be more friendly to serendipitous discoveries.^[97]

The "replication crisis" is compounded by a finding, published in a study summarized in 2021 by historian of science Naomi Oreskes, that nonreplicable studies are cited oftener than replicable ones: in other words, that bad science seems to get more attention than good science. If a substantial proportion of science is unreplicable, it will not provide a valid basis for decision-making and may delay the use of science for developing new medicines and technologies. It may also undermine the public's trust, making it harder to get people vaccinated or act against climate change.^[98]

The study tracked papers – in psychology journals, economics journals, and in Science and Nature – with documented failures of replication. The unreplicable papers were cited more than average, even after news of their unreplicability had been published.^[98]

"These results," writes Oreskes, "parallel those of a 2018 study. An analysis of 126,000 rumor cascades on Twitter showed that false news spread faster and reached more people than verified true claims. [I]t was people, not [ro]bots, who were responsible for the disproportionate spread of falsehoods online."^[98]

A 2016 Scientific American report highlights the role of rediscovery in science. Indiana University Bloomington researchers combed through 22 million scientific papers published over the previous century and found dozens of "Sleeping Beauties"—studies that lay dormant for years before getting noticed.^[99] The top finds, which languished longest and later received the most intense attention from scientists, came from the fields of chemistry, physics, and statistics. The dormant findings were wakened by scientists from other disciplines, such as medicine, in search of fresh insights, and by the ability to test once-theoretical postulations.^[99] Sleeping Beauties will likely become even more common in the future because of increasing accessibility of scientific literature.^[99] The Scientific American report lists the top 15 Sleeping Beauties: 7 in chemistry, 5 in physics, 2 in statistics, and 1 in metallurgy.^[99] Examples include:

Herbert Freundlich's "Concerning Adsorption in Solutions" (1906), the first mathematical model of adsorption, when atoms or molecules adhere to a surface. Today both environmental remediation and decontamination in industrial settings rely heavily on adsorption.^[99]

A. Einstein, B. Podolsky and N. Rosen, "Can Quantum-Mechanical Description of Physical Reality Be Considered Complete?" Physical Review, vol. 47 (May 15, 1935), pp. 777–780. This famous thought experiment in quantum physics—now known as the EPR paradox, after the authors' surname initials—was discussed theoretically when it first came out. It was not until the 1970s that physics had the experimental means to test quantum entanglement.^[99]

J[ohn] Turkevich, P. C. Stevenson, J. Hillier, "A Study of the Nucleation and Growth Processes in the Synthesis of Colloidal Gold", Discuss. Faraday. Soc., 1951, 11, pp. 55–75, explains how to suspend gold nanoparticles in liquid. It owes its awakening to medicine, which now employs gold nanoparticles to detect tumors and deliver drugs.^[99]

William S. Hummers and Richard E Offeman, "Preparation of Graphitic Oxide", Journal of the American Chemical Society, vol. 80, no. 6 (March 20, 1958), p. 1339, introduced Hummers' Method, a technique for making graphite oxide. Recent interest in graphene's potential has brought the 1958 paper to attention. Graphite oxide could serve as a reliable intermediate for the 2-D material.^[99]

Historians and sociologists have remarked the occurrence, in science, of "multiple independent discovery". Sociologist Robert K. Merton defined such "multiples" as instances in which similar discoveries are made by scientists working independently of each other.^[100] "Sometimes the discoveries are simultaneous or almost so; sometimes a scientist will make a new discovery which, unknown to him, somebody else has made years before."^[101][102] Commonly cited examples of multiple independent discovery are the 17th-century independent formulation of calculus by Isaac Newton, Gottfried Wilhelm Leibniz, and others;^[103] the 18th-century independent discovery of oxygen by Carl Wilhelm Scheele, Joseph Priestley, Antoine Lavoisier, and others; and the 19th-century independent formulation of the theory of evolution of species by Charles Darwin and Alfred Russel Wallace.^[104]

Merton contrasted a "multiple" with a "singleton" — a discovery that has been made uniquely by a single scientist or group of scientists working together.^[105] He believed that it is multiple discoveries, rather than unique ones, that represent the common pattern in science.^[106]

Multiple discoveries in the history of science provide evidence for evolutionary models of science and technology, such as memetics (the study of self-replicating units of culture), evolutionary epistemology (which applies the concepts of biological evolution to study of the growth of human knowledge), and cultural selection theory (which studies sociological and cultural evolution in a Darwinian manner). A recombinant-DNA-inspired "paradigm of paradigms", describing a mechanism of "recombinant conceptualization", predicates that a new concept arises through the crossing of pre-existing concepts and facts. This is what is meant when one says that a scientist, scholar, or artist has been "influenced by" another — etymologically, that a concept of the latter's has "flowed into" the mind of the former.^[107]

The phenomenon of multiple independent discoveries and inventions can be viewed as a consequence of Bolesław Prus' three laws of gradualness, dependence, and combination (see "Discoveries and inventions", above). The first two laws may, in turn, be seen as corollaries to the third law, since the laws of gradualness and dependence imply the impossibility of certain scientific or technological advances pending the availability of certain theories, facts, or technologies that must be combined to produce a given scientific or technological advance.

Technology – the application of discoveries to practical matters – showed a remarkable acceleration in what economist Robert J. Gordon has identified as "the special century" that spanned the period up to 1970. By then, he writes, all the key technologies of modern life were in place: sanitation, electricity, mechanized agriculture, highways, air travel, telecommunications, and the like. The one signature technology of the 21st century has been the iPhone. Meanwhile, a long list of much-publicized potential major technologies remain in the prototype phase, including self-driving cars, flying cars, augmented-reality glasses, gene therapy, and nuclear fusion. An urgent goal for the 21st century, writes Gordon, is to undo some of the consequences of the last great technology boom by developing affordable zero- and negative-emissions technologies.^[108]

Technology is the sum of techniques, skills, methods, and processes used in the production of goods or services or in the accomplishment of objectives, such as scientific investigation. Paradoxically, technology, so conceived, has sometimes been noted to take primacy over the ends themselves – even to their detriment. Laura Grego and David Wright, writing in 2019 in Scientific American, observe that "Current U.S. missile defense plans are being driven largely by technology, politics and fear. Missile defenses will not allow us to escape our vulnerability to nuclear weapons. Instead large-scale developments will create barriers to taking real steps toward reducing nuclear risks—by blocking further cuts in nuclear arsenals and potentially spurring new deployments."^[109]

Yale University physicist-astronomer Priyamvada Natarajan, writing of the virtually-simultaneous 1846 discovery of the planet Neptune by Urbain Le Verrier and John Couch Adams (after other astronomers, as early as Galileo Galilei in 1612, had unwittingly observed the planet), comments:

The episode is but one of many that proves science is not a dispassionate, neutral, and objective endeavor but rather one in which the violent clash of ideas and personal ambitions often combines with serendipity to propel new discoveries.^[110]

A practical question concerns the traits that enable some individuals to achieve extraordinary results in their fields of work—and how such creativity can be fostered. Melissa Schilling, a student of innovation strategy, has identified some traits shared by eight major innovators in natural science or technology: Benjamin Franklin (1706–90), Thomas Edison (1847–1931), Nikola Tesla (1856–1943), Maria Skłodowska Curie (1867–1934), Dean Kamen (born 1951), Steve Jobs (1955–2011), Albert Einstein (1879–1955), and Elon Musk (born 1971).^[111]

Schilling chose innovators in natural science and technology rather than in other fields because she found much more consensus about important contributions to natural science and technology than, for example, to art or music.^[112] She further limited the set to individuals associated with multiple innovations. "When an individual is associated with only a single major invention, it is much harder to know whether the invention was caused by the inventor's personal characteristics or by simply being at the right place at the right time."^[113]

The eight individuals were all extremely intelligent, but "that is not enough to make someone a serial breakthrough innovator."^[111] Nearly all these innovators showed very high levels of social detachment, or separateness (a notable exception being Benjamin Franklin).^[114] "Their isolation meant that they were less exposed to dominant ideas and norms, and their sense of not belonging meant that even when exposed to dominant ideas and norms, they were often less inclined to adopt them."^[115] From an early age, they had all shown extreme faith in their ability to overcome obstacles—what psychology calls "self-efficacy".^[115]

"Most [of them, writes Schilling] were driven by idealism, a superordinate goal that was more important than their own comfort, reputation, or families. Nikola Tesla wanted to free mankind from labor through unlimited free energy and to achieve international peace through global communication. Elon Musk wants to solve the world's energy problems and colonize Mars. Benjamin Franklin was seeking greater social harmony and productivity through the ideals of egalitarianism, tolerance, industriousness, temperance, and charity. Marie Curie had been inspired by Polish Positivism's argument that Poland, which was under Tsarist Russian rule, could be preserved only through the pursuit of education and technological advance by all Poles—including women."^[116]

Most of the innovators also worked hard and tirelessly because they found work extremely rewarding. Some had an extremely high need for achievement. Many also appeared to find work autotelic—rewarding for its own sake.^[117] A surprisingly large portion of the breakthrough innovators have been autodidacts—self-taught persons—and excelled much more outside the classroom than inside.^[118]

"Almost all breakthrough innovation," writes Schilling, "starts with an unusual idea or with beliefs that break with conventional wisdom.... However, creative ideas alone are almost never enough. Many people have creative ideas, even brilliant ones. But usually we lack the time, knowledge, money, or motivation to act on those ideas." It is generally hard to get others' help in implementing original ideas because the ideas are often initially hard for others to understand and value. Thus each of Schilling's breakthrough innovators showed extraordinary effort and persistence.^[119] Even so, writes Schilling, "being at the right place at the right time still matter[ed]."^[120]

When Swiss botanist Simon Schwendener discovered in the 1860s that lichens were a symbiotic partnership between a fungus and an alga, his finding at first met with resistance from the scientific community. After his discovery that the fungus—which cannot make its own food—provides the lichen's structure, while the alga's contribution is its photosynthetic production of food, it was found that in some lichens a cyanobacterium provides the food—and a handful of lichen species contain both an alga and a cyanobacterium, along with the fungus.^[121]

A self-taught naturalist, Trevor Goward, has helped create a paradigm shift in the study of lichens and perhaps of all life-forms by doing something that people did in pre-scientific times: going out into nature and closely observing. His essays about lichens were largely ignored by most researchers because Goward has no scientific degrees and because some of his radical ideas are not supported by rigorous data.^[122]

When Goward told Toby Spribille, who at the time lacked a high-school education, about some of his lichenological ideas, Goward recalls, "He said I was delusional." Ultimately Spribille passed a high-school equivalency examination, obtained a Ph.D. in lichenology at the University of Graz in Austria, and became an assistant professor of the ecology and evolution of symbiosis at the University of Alberta. In July 2016 Spribille and his co-authors published a ground-breaking paper in Science revealing that many lichens contain a second fungus.

Spribille credits Goward with having "a huge influence on my thinking. [His essays] gave me license to think about lichens in [an unorthodox way] and freed me to see the patterns I worked out in Bryoria with my co-authors." Even so, "one of the most difficult things was allowing myself to have an open mind to the idea that 150 years of literature may have entirely missed the theoretical possibility that there would be more than one fungal partner in the lichen symbiosis." Spribille says that academia's emphasis on the canon of what others have established as important is inherently limiting.^[123]

Contrary to previous studies indicating that higher intelligence makes for better leaders in various fields of endeavor, later research suggests that, at a certain point, a higher IQ can be viewed as harmful.^[124] Decades ago, psychologist Dean Simonton suggested that brilliant leaders' words may go over people's heads, their solutions could be more complicated to implement, and followers might find it harder to relate to them. At last, in the July 2017 Journal of Applied Psychology, he and two colleagues published the results of actual tests of the hypothesis.^[124][125]

Studied were 379 men and women business leaders in 30 countries, including the fields of banking, retail, and technology. The managers took IQ tests—an imperfect but robust predictor of performance in many areas—and each was rated on leadership style and effectiveness by an average of 8 co-workers. IQ correlated positively with ratings of leadership effectiveness, strategy formation, vision, and several other characteristics—up to a point. The ratings peaked at an IQ of about 120, which is higher than some 80% of office workers. Beyond that, the ratings declined. The researchers suggested that the ideal IQ could be higher or lower in various fields, depending on whether technical or social skills are more valued in a given work culture.^[124]

Psychologist Paul Sackett, not involved in the research, comments: "To me, the right interpretation of the work would be that it highlights a need to understand what high-IQ leaders do that leads to lower perceptions by followers. The wrong interpretation would be,'Don't hire high-IQ leaders.'"^[124] The study's lead author, psychologist John Antonakis, suggests that leaders should use their intelligence to generate creative metaphors that will persuade and inspire others. "I think the only way a smart person can signal their intelligence appropriately and still connect with the people," says Antonakis, "is to speak in charismatic ways."^[124]

Academic specialization produces great benefits for science and technology by focusing effort on discrete disciplines. But excessively narrow specialization can act as a roadblock to productive collaboration between traditional disciplines.

In 2017, in Manhattan, James Harris Simons, a noted mathematician and retired founder of one of the world's largest hedge funds, inaugurated the Flatiron Institute, a nonprofit enterprise whose goal is to apply his hedge fund's analytical strategies to projects dedicated to expanding knowledge and helping humanity.^[126] He has established computational divisions for research in astrophysics, biology, and quantum physics,^[127] and an interdisciplinary division for climate modelling that interfaces geology, oceanography, atmospheric science, biology, and climatology.^[128]

The latter, fourth Flatiron Institute division was inspired by a 2017 presentation to the institute's leadership by John Grotzinger, a "bio-geoscientist" from the California Institute of Technology, who explained the challenges of climate modelling. Grotzinger was a specialist in historical climate change—specifically, what had caused the great Permian extinction, during which virtually all species died. To properly assess this cataclysm, one had to understand both the rock record and the ocean's composition, but geologists did not interact much with physical oceanographers. Grotzinger's own best collaboration had resulted from a fortuitous lunch with an oceanographer. Climate modelling was an intrinsically difficult problem made worse by academia's structural divisions. "If you had it all under one umbrella... it could result [much sooner] in a major breakthrough." Simons and his team found Grotzinger's presentation compelling, and the Flatiron Institute decided to establish its fourth and final computational division.^[128]

Sociologist Harriet Zuckerman, in her 1977 study of natural-science Nobel laureates in the United States, was struck by the fact that more than half (48) of the 92 laureates who did their prize-winning research in the U.S. by 1972 had worked either as students, postdoctorates, or junior collaborators under older Nobel laureates. Furthermore, those 48 future laureates had worked under a total of 71 laureate masters.^[129][o]

Social viscosity ensures that not every qualified novice scientist attains access to the most productive centers of scientific thought. Nevertheless, writes Zuckerman, "To some extent, students of promise can choose masters with whom to work and masters can choose among the cohorts of students who present themselves for study. This process of bilateral assortative selection is conspicuously at work among the ultra-elite of science. Actual and prospective members of that elite select their scientist parents and therewith their scientist ancestors just as later they select their scientist progeny and therewith their scientist descendants."^[131]

Zuckerman writes: "[T]he lines of elite apprentices to elite masters who had themselves been elite apprentices, and so on indefinitely, often reach far back into the history of science, long before 1900, when [Alfred] Nobel's will inaugurated what now amounts to the International Academy of Sciences. As an example of the many long historical chains of elite masters and apprentices, consider the German-born English laureate Hans Krebs (1953), who traces his scientific lineage [...] back through his master, the 1931 laureate Otto Warburg. Warburg had studied with Emil Fis[c]her [1852–1919], recipient of a prize in 1902 at the age of 50, three years before it was awarded [in 1905] to his teacher, Adolf von Baeyer [1835–1917], at age 70. This lineage of four Nobel masters and apprentices has its own pre-Nobelian antecedents. Von Baeyer had been the apprentice of F[riedrich] A[ugust] Kekulé [1829–1896], whose ideas of structural formulae revolutionized organic chemistry and who is perhaps best known for the often retold story about his having hit upon the ring structure of benzene in a dream (1865). Kekulé himself had been trained by the great organic chemist Justus von Liebig (1803–1873), who had studied at the Sorbonne with the master J[oseph] L[ouis] Gay-Lussac (1778–1850), himself once apprenticed to Claude Louis Berthollet (1748–1822). Among his many institutional and cognitive accomplishments, Berthollet helped found the École Polytechnique, served as science advisor to Napoleon in Egypt, and, more significant for our purposes here, worked with [Antoine] Lavoisier [1743–1794] to revise the standard system of chemical nomenclature."^[132]

Sociologist Michael P. Farrell has studied close creative groups and writes: "Most of the fragile insights that laid the foundation of a new vision emerged not when the whole group was together, and not when members worked alone, but when they collaborated and repsonded to one another in pairs."^[133] François Jacob, who, with Jacques Monod, pioneered the study of gene regulation, notes that by the mid-20th century, most research in molecular biology was conducted by twosomes. "Two are better than one for dreaming up theories and constructing models," writes Jacob. "For with two minds working on a problem, ideas fly thicker and faster. They are bounced from partner to partner.... And in the process, illusions are sooner nipped in the bud." As of 2018, in the previous 35 years, some half of Nobel Prizes in Physiology or Medicine had gone to scientific partnerships.^[134] James Somers describes a remarkable partnership between Google's top software engineers, Jeff Dean and Sanjay Ghemawat.^[135]

Twosome collaborations have also been prominent in creative endeavors outside the natural sciences and technology; examples are Claude Monet's and Pierre-Auguste Renoir's 1869 joint creation of Impressionism, Pablo Picasso's and Georges Braque's six-year collaborative creation of Cubism, and John Lennon's and Paul McCartney's collaborations on Beatles songs. "Everyone", writes James Somers, "falls into creative ruts, but two people rarely do so at the same time."^[136]

The same point was made by Francis Crick, member of a famous scientific duo, Francis Crick and James Watson, who together discovered the structure of the genetic material, DNA. At the end of a PBS television documentary on James Watson, in a video clipping Crick explains to Watson that their collaboration had been crucial to their discovery because, when one of them was wrong, the other would set him straight.^[137]

What has been dubbed "Big Science" emerged from the United States' World War II Manhattan Project that produced the world's first nuclear weapons; and Big Science has since been associated with physics, which requires massive particle accelerators. In biology, Big Science debuted in 1990 with the Human Genome Project to sequence human DNA. In 2013 neuroscience became a Big Science domain when the U.S. announced a BRAIN Initiative and the European Union announced a Human Brain Project. Major new brain-research initiatives were also announced by Israel, Canada, Australia, New Zealand, Japan, and China.^[138]

Earlier successful Big Science projects had habituated politicians, mass media, and the public to view Big Science programs with sometimes uncritical favor.^[139]

The U.S.'s BRAIN Initiative was inspired by concern about the spread and cost of mental disorders and by excitement about new brain-manipulation technologies such as optogenetics.^[140] After some early false starts, the U.S. National Institute of Mental Health let the country's brain scientists define the BRAIN Initiative, and this led to an ambitious interdisciplinary program to develop new technological tools to better monitor, measure, and simulate the brain. Competition in research was ensured by the National Institute of Mental Health's peer-review process.^[139]

In the European Union, the European Commission's Human Brain Project got off to a rockier start because political and economic considerations obscured questions concerning the feasibility of the Project's initial scientific program, based principally on computer modeling of neural circuits. Four years earlier, in 2009, fearing that the European Union would fall further behind the U.S. in computer and other technologies, the European Union had begun creating a competition for Big Science projects, and the initial program for the Human Brain Project seemed a good fit for a European program that might take a lead in advanced and emerging technologies.^[140] Only in 2015, after over 800 European neuroscientists threatened to boycott the European-wide collaboration, were changes introduced into the Human Brain Project, supplanting many of the original political and economic considerations with scientific ones.^[141]

As of 2019, the European Union's Human Brain Project had not lived up to its extravagant promise.^[142]

Nathan Myhrvold, former Microsoft chief technology officer and founder of Microsoft Research, argues that the funding of basic science cannot be left to the private sector—that "without government resources, basic science will grind to a halt."^[143] He notes that Albert Einstein's general theory of relativity, published in 1915, did not spring full-blown from his brain in a eureka moment; he worked at it for years—finally driven to complete it by a rivalry with mathematician David Hilbert.^[143] The history of almost any iconic scientific discovery or technological invention—the lightbulb, the transistor, DNA, even the Internet—shows that the famous names credited with the breakthrough "were only a few steps ahead of a pack of competitors." Some writers and elected officials have used this phenomenon of "parallel innovation" to argue against public financing of basic research: government, they assert, should leave it to companies to finance the research they need.^[143]

Myhrvold writes that such arguments are dangerously wrong: without government support, most basic scientific research will never happen. "This is most clearly true for the kind of pure research that has delivered... great intellectual benefits but no profits, such as the work that brought us the Higgs boson, or the understanding that a supermassive black hole sits at the center of the Milky Way, or the discovery of methane seas on the surface of Saturn's moon Titan. Company research laboratories used to do this kind of work: experimental evidence for the Big Bang was discovered at AT&T's Bell Labs, resulting in a Nobel Prize. Now those days are gone."^[143]

Even in applied fields such as materials science and computer science, writes Myhrvold, "companies now understand that basic research is a form of charity—so they avoid it." Bell Labs scientists created the transistor, but that invention earned billions for Intel and Microsoft. Xerox PARC engineers invented the modern graphical user interface, but Apple and Microsoft profited most. IBM researchers pioneered the use of giant magnetoresistance to boost hard-disk capacity but soon lost the disk-drive business to Seagate and Western Digital.^[143]

Company researchers now have to focus narrowly on innovations that can quickly bring revenue; otherwise the research budget could not be justified to the company's investors. "Those who believe profit-driven companies will altruistically pay for basic science that has wide-ranging benefits—but mostly to others and not for a generation—are naive.... If government were to leave it to the private sector to pay for basic research, most science would come to a screeching halt. What research survived would be done largely in secret, for fear of handing the next big thing to a rival."^[143]

Governmental investment is equally vital in the field of biological research. According to William A. Haseltine, a former Harvard Medical School professor and founder of that university's cancer and HIV / AIDS research departments, early efforts to control the COVID-19 pandemic were hampered by governments and industry everywhere having "pulled the plug on coronavirus research funding in 2006 after the first SARS [...] pandemic faded away and again in the years immediately following the MERS [outbreak, also caused by a coronavirus] when it seemed to be controllable.^[144] [...] The development of promising anti-SARS and MERS drugs, which might have been active against SARS–CoV-2 [in the Covid-19 pandemic] as well, was left unfinished for lack of money."^[145] Haseltine continues:

We learned from the HIV crisis that it was important to have research pipelines already established. [It was c]ancer research in the 1950s, 1960s and 1970s [that] built a foundation for HIV / Aids studies. [During those decades t]he government [had] responded to public concerns, sharply increasing federal funding of cancer research [...]. These efforts [had] culminated in Congress's approval of President Richard Nixon's National Cancer Act in 1971. This [had] built the science we needed to identify and understand HIV in the 1980s, although of course no one knew that payoff was coming.^[145]

In the 1980s the Reagan administration did not want to talk about AIDS or commit much funding to HIV research. [But o]nce the news broke that actor Rock Hudson was seriously ill with AIDS, [...] $320 million [were added to] the fiscal 1986 budget for AIDS research. [...] I helped [...] design this first congressionally funded AIDS research program with Anthony Fauci, the doctor now leading [the U.S.] fight against COVID-19.^[145] [...]

[The] tool set for virus and pharmaceutical research has improved enormously in the past 36 years since HIV was discovered. What used to take five or 10 years in the 1980s and 1990s in many cases now can be done in five or 10 months. We can rapidly identify and synthesize chemicals to predict which drugs will be effective. We can do cryoelectron microscopy to probe virus structures and simulate molecule-by-molecule interactions in a matter of weeks – something that used to take years. The lesson is to never let down our guard when it comes to funding antiviral research. We would have no hope of beating COVID-19 if it were not for the molecular biology gains we made during earlier virus battles. What we learn this time around will help us [...] during the next pandemic, but we must keep the money coming.^[145]

A complementary perspective on the funding of scientific research is given by D.T. Max, writing about the Flatiron Institute, a computational center set up in 2017 in Manhattan to provide scientists with mathematical assistance. The Flatiron Institute was established by James Harris Simons, a mathematician who had used mathematical algorithms to make himself a Wall Street billionaire. The institute has three computational divisions dedicated respectively to astrophysics, biology, and quantum physics, and is working on a fourth division for climate modeling that will involve interfaces of geology, oceanography, atmospheric science, biology, and climatology.^[128]

The Flatiron Institute is part of a trend in the sciences toward privately funded research. In the United States, basic science has traditionally been financed by universities or the government, but private institutes are often faster and more focused. Since the 1990s, when Silicon Valley began producing billionaires, private institutes have sprung up across the U.S. In 1997 Larry Ellison launched the Ellison Medical Foundation to study the biology of aging. In 2003 Paul Allen founded the Allen Institute for Brain Science. In 2010 Eric Schmidt founded the Schmidt Ocean Institute.^[146]

These institutes have done much good, partly by providing alternatives to more rigid systems. But private foundations also have liabilities. Wealthy benefactors tend to direct their funding toward their personal enthusiasms. And foundations are not taxed; much of the money that supports them would otherwise have gone to the government.^[146]

John P.A. Ioannidis, of Stanford University Medical School, writes that "There is increasing evidence that some of the ways we conduct, evaluate, report and disseminate research are miserably ineffective. A series of papers in 2014 in The Lancet... estimated that 85 percent of investment in biomedical research is wasted. Many other disciplines have similar problems."^[147] Ioannidis identifies some science-funding biases that undermine the efficiency of the scientific enterprise, and proposes solutions:

Funding too few scientists: "[M]ajor success [in scientific research] is largely the result of luck, as well as hard work. The investigators currently enjoying huge funding are not necessarily genuine superstars; they may simply be the best connected." Solutions: "Use a lottery to decide which grant applications to fund (perhaps after they pass a basic review).... Shift... funds from senior people to younger researchers..."^[147]

No reward for transparency: "Many scientific protocols, analysis methods, computational processes and data are opaque. [M]any top findings cannot be reproduced. That is the case for two out of three top psychology papers, one out of three top papers in experimental economics and more than 75 percent of top papers identifying new cancer drug targets. [S]cientists are not rewarded for sharing their techniques." Solutions: "Create better infrastructure for enabling transparency, openness and sharing. Make transparency a prerequisite for funding. [P]referentially hire, promote or tenure... champions of transparency."^[147]

No encouragement for replication: Replication is indispensable to the scientific method. Yet, under pressure to produce new discoveries, researchers tend to have little incentive, and much counterincentive, to try replicating results of previous studies. Solutions: "Funding agencies must pay for replication studies. Scientists' advancement should be based not only on their discoveries but also on their replication track record."^[147]

No funding for young scientists: "Werner Heisenberg, Albert Einstein, Paul Dirac and Wolfgang Pauli made their top contributions in their mid-20s." But the average age of biomedical scientists receiving their first substantial grant is 46. The average age for a full professor in the U.S. is 55. Solutions: "A larger proportion of funding should be earmarked for young investigators. Universities should try to shift the aging distribution of their faculty by hiring more young investigators."^[147]

Biased funding sources: "Most funding for research and development in the U.S. comes not from the government but from private, for-profit sources, raising unavoidable conflicts of interest and pressure to deliver results favorable to the sponsor." Solutions: "Restrict or even ban funding that has overt conflicts of interest. Journals should not accept research with such conflicts. For less conspicuous conflicts, at a minimum ensure transparent and thorough disclosure."^[148][p]

Funding the wrong fields: "Well-funded fields attract more scientists to work for them, which increases their lobbying reach, fueling a vicious circle. Some entrenched fields absorb enormous funding even though they have clearly demonstrated limited yield or uncorrectable flaws." Solutions: "Independent, impartial assessment of output is necessary for lavishly funded fields. More funds should be earmarked for new fields and fields that are high risk. Researchers should be encouraged to switch fields, whereas currently they are incentivized to focus in one area."^[148]

Not spending enough: The U.S. military budget ($886 billion) is 24 times the budget of the National Institutes of Health ($37 billion). "Investment in science benefits society at large, yet attempts to convince the public often make matters worse when otherwise well-intentioned science leaders promise the impossible, such as promptly eliminating all cancer or Alzheimer's disease." Solutions: "We need to communicate how science funding is used by making the process of science clearer, including the number of scientists it takes to make major accomplishments.... We would also make a more convincing case for science if we could show that we do work hard on improving how we run it."^[148]

Rewarding big spenders: "Hiring, promotion and tenure decisions primarily rest on a researcher's ability to secure high levels of funding. But the expense of a project does not necessarily correlate with its importance. Such reward structures select mostly for politically savvy managers who know how to absorb money." Solutions: "We should reward scientists for high-quality work, reproducibility and social value rather than for securing funding. Excellent research can be done with little to no funding other than protected time. Institutions should provide this time and respect scientists who can do great work without wasting tons of money."^[148]

No funding for high-risk ideas: "The pressure that taxpayer money be 'well spent' leads government funders to back projects most likely to pay off with a positive result, even if riskier projects might lead to more important, but less assured, advances. Industry also avoids investing in high-risk projects... Innovation is extremely difficult, if not impossible, to predict..." Solutions: "Fund excellent scientists rather than projects and give them freedom to pursue research avenues as they see fit. Some institutions such as Howard Hughes Medical Institute already use this model with success." It must be communicated to the public and to policy-makers that science is a cumulative investment, that no one can know in advance which projects will succeed, and that success must be judged on the total agenda, not on a single experiment or result.^[148]

Lack of good data: "There is relatively limited evidence about which scientific practices work best. We need more research on research ('meta-research') to understand how to best perform, evaluate, review, disseminate and reward science." Solutions: "We should invest in studying how to get the best science and how to choose and reward the best scientists."^[148]

Naomi Oreskes, professor of the history of science at Harvard University, writes about the desirability of diversity in the backgrounds of scientists.

The history of science is rife with [...] cases of misogyny, prejudice and bias. For centuries biologists promoted false theories of female inferiority, and scientific institutions typically barred women's participation. Historian of science [...] Margaret Rossiter has documented how, in the mid-19th century, female scientists created their own scientific societies to compensate for their male colleagues' refusal to acknowledge their work. Sharon Bertsch McGrayne filled an entire volume with the stories of women who should have been awarded the Nobel Prize for work that they did in collaboration with male colleagues – or, worse, that they had stolen by them. [...] Racial bias has been at least as pernicious as gender bias; it was scientists, after all, who codified the concept of race as a biological category that was not simply descriptive but also hierarchical.^[150]

[...] [C]ognitive science shows that humans are prone to bias, misperception, motivated reasoning and other intellectual pitfalls. Because reasoning is slow and difficult, we rely on heuristics – intellectual shortcuts that often work but sometimes fail spectacularly. (Believing that men are, in general, better than women in math is one tiring example.) [...]^[150]

[...] Science is a collective effort, and it works best when scientific communities are diverse. [H]eterogeneous communities are more likely than homogeneous ones to be able to identify blind spots and correct them. Science does not correct itself; scientists correct one another through critical interrogation. And that means being willing to interrogate not just claims about the external world but claims about [scientists'] own practices and processes as well.^[150]

Claire Pomeroy, president of the Lasker Foundation, which is dedicated to advancing medical research, points out that women scientists continue to be subjected to discrimination in professional advancement.^[151]

Though the percentage of doctorates awarded to women in life sciences in the United States increased from 15 to 52 percent between 1969 and 2009, only a third of assistant professors and less than a fifth of full professors in biology-related fields in 2009 were women. Women make up only 15 percent of permanent department chairs in medical schools and barely 16 percent of medical-school deans.^[151]

The problem is a culture of unconscious bias that leaves many women feeling demoralized and marginalized. In one study, science faculty were given identical résumés in which the names and genders of two applicants were interchanged; both male and female faculty judged the male applicant to be more competent and offered him a higher salary.^[151]

Unconscious bias also appears as "microassaults" against women scientists: purportedly insignificant sexist jokes and insults that accumulate over the years and undermine confidence and ambition. Writes Claire Pomeroy: "Each time it is assumed that the only woman in the lab group will play the role of recording secretary, each time a research plan becomes finalized in the men's lavatory between conference sessions, each time a woman is not invited to go out for a beer after the plenary lecture to talk shop, the damage is reinforced."^[151]

"When I speak to groups of women scientists," writes Pomeroy, "I often ask them if they have ever been in a meeting where they made a recommendation, had it ignored, and then heard a man receive praise and support for making the same point a few minutes later. Each time the majority of women in the audience raise their hands. Microassaults are especially damaging when they come from a high-school science teacher, college mentor, university dean or a member of the scientific elite who has been awarded a prestigious prize—the very people who should be inspiring and supporting the next generation of scientists."^[151]

Sexual harassment is more prevalent in academia than in any other social sector except the military. A June 2018 report by the National Academies of Sciences, Engineering, and Medicine states that sexual harassment hurts individuals, diminishes the pool of scientific talent, and ultimately damages the integrity of science.^[152]

Paula Johnson, co-chair of the committee that drew up the report, describes some measures for preventing sexual harassment in science. One would be to replace trainees' individual mentoring with group mentoring, and to uncouple the mentoring relationship from the trainee's financial dependence on the mentor. Another way would be to prohibit the use of confidentiality agreements in connection with harassment cases.^[152]

A novel approach to the reporting of sexual harassment, dubbed Callisto, that has been adopted by some institutions of higher education, lets aggrieved persons record experiences of sexual harassment, date-stamped, without actually formally reporting them. This program lets people see if others have recorded experiences of harassment from the same individual, and share information anonymously.^[152]

Psychologist Andrei Cimpian and philosophy professor Sarah-Jane Leslie have proposed a theory to explain why American women and African-Americans are often subtly deterred from seeking to enter certain academic fields by a misplaced emphasis on genius.^[153] Cimpian and Leslie had noticed that their respective fields are similar in their substance but hold different views on what is important for success. Much more than psychologists, philosophers value a certain kind of person: the "brilliant superstar" with an exceptional mind. Psychologists are more likely to believe that the leading lights in psychology grew to achieve their positions through hard work and experience.^[154] In 2015, women accounted for less than 30% of doctorates granted in philosophy; African-Americans made up only 1% of philosophy Ph.D.s. Psychology, on the other hand, has been successful in attracting women (72% of 2015 psychology Ph.D.s) and African-Americans (6% of psychology Ph.D.s).^[155]

An early insight into these disparities was provided to Cimpian and Leslie by the work of psychologist Carol Dweck. She and her colleagues had shown that a person's beliefs about ability matter a great deal for that person's ultimate success. A person who sees talent as a stable trait is motivated to "show off this aptitude" and to avoid making mistakes. By contrast, a person who adopts a "growth mindset" sees his or her current capacity as a work in progress: for such a person, mistakes are not an indictment but a valuable signal highlighting which of their skills are in need of work.^[156] Cimpian and Leslie and their collaborators tested the hypothesis that attitudes, about "genius" and about the unacceptability of making mistakes, within various academic fields may account for the relative attractiveness of those fields for American women and African-Americans. They did so by contacting academic professionals from a wide range of disciplines and asking them whether they thought that some form of exceptional intellectual talent was required for success in their field. The answers received from almost 2,000 academics in 30 fields matched the distribution of Ph.D.s in the way that Cimpian and Leslie had expected: fields that placed more value on brilliance also conferred fewer Ph.D.s on women and African-Americans. The proportion of women and African-American Ph.D.s in psychology, for example, was higher than the parallel proportions for philosophy, mathematics, or physics.^[157]

Further investigation showed that non-academics share similar ideas of which fields require brilliance. Exposure to these ideas at home or school could discourage young members of stereotyped groups from pursuing certain careers, such as those in the natural sciences or engineering. To explore this, Cimpian and Leslie asked hundreds of five-, six-, and seven-year-old boys and girls questions that measured whether they associated being "really, really smart" (i.e., "brilliant") with their sex. The results, published in January 2017 in Science, were consistent with scientific literature on the early acquisition of sex stereotypes. Five-year-old boys and girls showed no difference in their self-assessment; but by age six, girls were less likely to think that girls are "really, really smart." The authors next introduced another group of five-, six-, and seven-year-olds to unfamiliar gamelike activities that the authors described as being "for children who are really, really smart." Comparison of boys' and girls' interest in these activities at each age showed no sex difference at age five but significantly greater interest from boys at ages six and seven—exactly the ages when stereotypes emerge.^[158]

Cimpian and Leslie conclude that, "Given current societal stereotypes, messages that portray [genius or brilliance] as singularly necessary [for academic success] may needlessly discourage talented members of stereotyped groups."^[158]

Largely as a result of his growing popularity, astronomer and science popularizer Carl Sagan, creator of the 1980 PBS TV Cosmos series, came to be ridiculed by scientist peers and failed to receive tenure at Harvard University in the 1960s and membership in the National Academy of Sciences in the 1990s. The eponymous "Sagan effect" persists: as a group, scientists still discourage individual investigators from engaging with the public unless they are already well-established senior researchers.^[159][160]

The operation of the Sagan effect deprives society of the full range of expertise needed to make informed decisions about complex questions, including genetic engineering, climate change, and energy alternatives. Fewer scientific voices mean fewer arguments to counter antiscience or pseudoscientific discussion. The Sagan effect also creates the false impression that science is the domain of older white men (who dominate the senior ranks), thereby tending to discourage women and minorities from considering science careers.^[159]

A number of factors contribute to the Sagan effect's durability. At the height of the Scientific Revolution in the 17th century, many researchers emulated the example of Isaac Newton, who dedicated himself to physics and mathematics and never married. These scientists were viewed as pure seekers of truth who were not distracted by more mundane concerns. Similarly, today anything that takes scientists away from their research, such as having a hobby or taking part in public debates, can undermine their credibility as researchers.^[161]

Another, more prosaic factor in the Sagan effect's persistence may be professional jealousy.^[161]

However, there appear to be some signs that engaging with the rest of society is becoming less hazardous to a career in science. So many people have social-media accounts now that becoming a public figure is not as unusual for scientists as previously. Moreover, as traditional funding sources stagnate, going public sometimes leads to new, unconventional funding streams. A few institutions such as Emory University and the Massachusetts Institute of Technology may have begun to appreciate outreach as an area of academic activity, in addition to the traditional roles of research, teaching, and administration. Exceptional among federal funding agencies, the National Science Foundation now officially favors popularization.^[162][160]

Like infectious diseases, ideas in academia are contagious. But why some ideas gain great currency while equally good ones remain in relative obscurity had been unclear. A team of computer scientists has used an epidemiological model to simulate how ideas move from one academic institution to another. The model-based findings, published in October 2018, show that ideas originating at prestigious institutions cause bigger "epidemics" than equally good ideas from less prominent places. The finding reveals a big weakness in how science is done. Many highly trained people with good ideas do not obtain posts at the most prestigious institutions; much good work published by workers at less prestigious places is overlooked by other scientists and scholars because they are not paying attention.^[163]

Naomi Oreskes remarks on another drawback to deprecating public universities in favor of Ivy League schools: "In 1970 most jobs did not require a college degree. Today nearly all well-paying ones do. With the rise of artificial intelligence and the continued outsourcing of low-skilled and de-skilled jobs overseas, that trend most likely will accelerate. Those who care about equity of opportunity should pay less attention to the lucky few who get into Harvard or other highly selective private schools and more to public education, because for most Americans, the road to opportunity runs through public schools."^[164]

Resistance, among some of the public, to accepting vaccination and the reality of climate change may be traceable partly to several decades of partisan attacks on government, leading to distrust of government science and then of science generally.^[165]

Many scientists themselves have been loth to involve themselves in public policy debates for fear of losing credibility: they worry that if they participate in public debate on a contested question, they will be viewed as biased and discounted as partisan. However, studies show that most people want to hear from scientists on matters within their areas of expertise. Research also suggests that scientists can feel comfortable offering policy advice within their fields. "The ozone story", writes Naomi Oreskes, "is a case in point: no one knew better than ozone scientists about the cause of the dangerous hole and therefore what needed to be done to fix it."^[166]

Oreskes, however, identifies a factor that does "turn off" the public: scientists' frequent use of jargon – of expressions that tend to be misinterpreted by, or incomprehensible to, laypersons.^[165]

In climatological parlance, "positive feedback" refers to amplifying feedback loops, such as the ice-albedo feedback. ("Albedo", another piece of jargon, simply means "reflectivity".) The positive loop in question develops when global warming causes Arctic ice to melt, exposing water that is darker and reflects less of the sun's warming rays, leading to more warming, which leads to more melting... and so on. In climatology, such positive feedback is a bad thing; but for most laypersons, "it conjures reassuring images, such as receiving praise from your boss.".^[165]

When astronomers say "metals," they mean any element heavier than helium, which includes oxygen and nitrogen, a usage that is massively confusing not just to laypersons but also to chemists. [To astronomers] [t]he Big Dipper isn't a constellation [...] it is an "asterism" [...] In AI, there is machine "intelligence," which isn't intelligence at all but something more like "machine capability." In ecology, there are "ecosystem services," which you might reasonably think refers to companies that clean up oil spills, but it is [actually] ecological jargon for all the good things that the natural world does for us. [T]hen there's [...] the theory of "communication accommodation," which means speaking so that the listener can understand.^[165]

"[R]esearchers," writes Naomi Oreskes, "are often judged more by the quantity of their output than its quality. Universities [emphasize] metrics such as the numbers of published papers and citations when they make hiring, tenure and promotion decisions."^[167]

When – for a number of possible reasons – publication in legitimate peer-reviewed journals is not feasible, this often creates a perverse incentive to publish in "predatory journals", which do not uphold scientific standards. Some 8,000 such journals publish 420,000 papers annually – nearly a fifth of the scientific community's annual output of 2.5 million papers. The papers published in a predatory journal are listed in scientific databases alongside legitimate journals, making it hard to discern the difference.^[168]

One reason why some scientists publish in predatory journals is that prestigious scientific journals may charge scientists thousands of dollars for publishing, whereas a predatory journal typically charges less than $200. (Hence authors of papers in the predatory journals are disproportionately located in less wealthy countries and institutions.)^[169]

Publishing in predatory journals can be life-threatening when physicians and patients accept spurious claims about medical treatments; and invalid studies can wrongly influence public policy. More such predatory journals are appearing every year. In 2008 Jeffrey Beall, a University of Colorado librarian, developed a list of predatory journals which he updated for several years.^[170]

Naomi Oreskes argues that, "[t]o put an end to predatory practices, universities and other research institutions need to find ways to correct the incentives that lead scholars to prioritize publication quantity... Setting a maximum limit on the number of articles that hiring or funding committees can consider might help... as could placing less importance on the number of citations an author gets. After all, the purpose of science is not merely to produce papers. It is to produce papers that tell us something truthful and meaningful about the world."^[171]

The perverse incentive to "publish or perish" is often facilitated by the fabrication of data. A classic example is the identical-twin-studies results of Cyril Burt, which – soon after Burt's death – were found to have been based on fabricated data.

Writes Gideon Lewish-Kraus:

"One of the confounding things about the social sciences is that observational evidence can produce only correlations. [For example, t]o what extent is dishonesty [which is the subject of a number of social-science studies] a matter of character, and to what extent a matter of situation? Research misconduct is sometimes explained away by incentives – the publishing requirements for the job market, or the acclaim that can lead to consulting fees and Davos appearances. [...] The differences between p-hacking and fraud is one of degree. And once it becomes customary within a field to inflate results, the field selects for researchers inclined to do so."^[172]

Joe Simmons, a behavioral-science professor, writes:

"[A] field cannot reward truth if it does not or cannot decipher it, so it rewards other things instead. Interestingness. Novelty. Speed. Impact. Fantasy. And it effectively punishes the opposite. Intuitive Findings. Incremental Progress. Care. Curiosity. Reality."^[173]

Harvard University historian of science Naomi Oreskes writes that a theme at the 2024 World Economic Forum in Davos, Switzerland, was a "perceived need to 'accelerate breakthroughs in research and technology.'"^[174]

"[R]ecent years", however, writes Oreskes, "[have] seen important papers, written by prominent scientists and published in prestigious journals, retracted because of questionable data or methods." For example, the Davos meeting took place after the resignations – over questionably reliable academic papers – in 2023 of Stanford University president Marc Tessier-Lavigne and, in 2024, of Harvard University president Claudine Gay. "In one interesting case, Frances H. Arnold of the California Institute of Technology, who shared the 2018 Nobel Prize in Chemistry, voluntarily retracted a paper when her lab was unable to replicate her results – but after the paper had been published." Such incidents, suggests Oreskes, are likely to erode public trust in science and in experts generally.^[175]

Academics at leading universities in the United States and Europe are subject to perverse incentives to produce results – and lots of them – quickly. A study has put the number of papers published around 2023 by scientists and other scholars at over seven million annually, compared with less than a million in 1980. Another study found 265 authors – two-thirds in the medical and life sciences – who published on average a paper every five days.^[176]

"Good science [and scholarship take] time", writes Oreskes. "More than 50 years elapsed between the 1543 publication of Copernicus's magnum opus... and the broad scientific acceptance of the heliocentric model... Nearly a century passed between biochemist Friedrich Miescher's identification of the DNA molecule and suggestion that it might be involved in inheritance and the elucidation of its double-helix structure in the 1950s. And it took just about half a century for geologists and geophysicists to accept geophysicist Alfred Wegener's idea of continental drift."^[177]

• American Masters: Decoding Watson - PBS documentary about James Watson, co-discoverer of the structure of DNA, including interviews with Watson, his family, and colleagues. 2019-01-02.